Personal opinions are a prospective means to detect trends in products, services, or topics in the news. The Internet has led to the proliferation of online opinions that may be found in a variety of web-accessible sources. Online opinions may be found in reviews, ratings, and recommendations, which may be part of blogs, social networks, product web sites, news articles, online forums and so forth. The online opinions are useful to get anonymously consumers' insight into the latest trends which may be used to initiate future developments of new products and services, to improve the marketability of a product or service, to introduce a product in the marketplace, to reveal issues that affect a customer's opinion and to identify popular topics of interest.
Natural language programming tools, such as sentiment analysis, have attempted to automate the process of detecting opinions from online sources. These tools attempt to capture the sentiment expressed in sentences based on the use of particular words or phrases in a particular context. However, there are nuances with a human language, such as cultural factors, that make sentiment analysis unreliable for automatically detecting opinions from online data.